Why is trunk 'laggy'?

[omegaamigo]

I have two Cisco 4006s, one in each of two comms rooms ("West" and "East"). Most of our servers and other kit (including those providing our internet link) is in "East": desktop PCs patched to 'East' get good network performance connecting to our local servers and to the internet. Desktop PCs patched to 'West', however, do not.

There is some connectivity, and shell sessions to the 'East' servers are just about usable (it's also feasible to control one switch from the other, across the trunk), but as soon as there is any level of traffic, it kind-of grinds to a halt (for example, it is just about possible to load a small text-only web page, but as soon as it hits an image, it starts crawling).

This network is under test at the moment, and only has three servers and a couple of desktops; most of them are idle most of the time, so there's almost no network traffic.

General setup:

- Port 1/2 on West to Port 1/2 on East is the main trunk, across 1000-Base-LX (single-mode fibre);

- All but a handful of ports are on vlan 1 - vlan 105 has been defined for a small subnet, but is not currently in use;

- I plan to use VMPS soon - it is configured, but not enabled;

Details:

WestComms 4006:

begin
!
# ***** NON-DEFAULT CONFIGURATION *****
!
!
#time: Sat Jun 4 2005, 11:31:25
!
#version 7.6(6)
!
!
#system web interface version(s)
!
#dot1x
set feature dot1x-radius-keepalive disable
!
#system
set system name  WestComms
!
#frame distribution method
set port channel all distribution mac both
!
#vtp
set vtp domain ceu
set vlan 1 name default type ethernet mtu 1500 said 100001 state active
set vlan 105 name Mauve type ethernet mtu 1500 said 100105 state active
set vlan 1002 name fddi-default type fddi mtu 1500 said 101002 state active
set vlan 1004 name fddinet-default type fddinet mtu 1500 said 101004 state active stp ieee
set vlan 1005 name trnet-default type trbrf mtu 1500 said 101005 state active stp ibm
set vlan 1003 name token-ring-default type trcrf mtu 1500 said 101003 state active mode srb aremaxhop 7 stemaxhop 7 backupcrf off
!
#ip
set interface sc0 1 10.7.7.1/255.255.255.0 10.7.7.255

set interface sl0 down
set interface me1 down
set ip alias default         0.0.0.0
!
#vmps
set vmps server 10.7.7.1 primary
set vmps downloadserver 10.7.7.3 vmps.dat
!
#dns
set ip dns server 10.0.0.2 primary
set ip dns domain our.domain.name
!
#set boot command
set boot config-register 0x2
set boot system flash bootflash:cat4000-k8.7-6-6.bin
!
#multicast filter
set igmp filter disable
!
#module 1 : 2-port 1000BaseX Supervisor
set port name       1/2  TrunkToEast
set trunk 1/2  on dot1q 1-1005,1025-4094
!
#module 2 : 48-port 10/100BaseTx Ethernet
!
#module 3 : 48-port 10/100BaseTx Ethernet
!
#module 4 : 48-port 10/100BaseTx Ethernet
!
#module 5 : 48-port 10/100BaseTx Ethernet
!
#module 6 : 24-port 10/100/1000 Ethernet
end

EastComms 4006:

begin
!
# ***** NON-DEFAULT CONFIGURATION *****
!
!
#time: Sat Jan 1 2000, 00:00:11
!
#version 7.6(6)
!
!
#system web interface version(s)
!
#dot1x
set feature dot1x-radius-keepalive disable
!
#system
set system name  EastComms
!
#frame distribution method
set port channel all distribution mac both
!
#vtp
set vtp domain ceu
set vtp mode client
!
#ip
set interface sc0 1 10.7.7.2/255.255.255.0 10.7.7.255

set interface sl0 down
set interface me1 down
set ip alias default         0.0.0.0
!
#dns
set ip dns server 10.0.0.2 primary
set ip dns domain our.domain.name
!
#set boot command
set boot config-register 0x2102
clear boot system all
!
#multicast filter
set igmp filter disable
!
#module 1 : 2-port 1000BaseX Supervisor
set port name       1/2  TrunkToWest
set trunk 1/2  on dot1q 1-1005,1025-4094
!
#module 2 : 48-port 10/100BaseTx Ethernet
!
#module 3 : 48-port 10/100BaseTx Ethernet
!
#module 4 : 48-port 10/100BaseTx Ethernet
!
#module 5 : 48-port 10/100BaseTx Ethernet
set vlan 105  5/25-48
!
#module 6 : 24-port 10/100/1000 Ethernet
end

Anyone have any suggestions or help here? I'm not sure how to proceed in troubleshooting this further ...

Thanks ...

 

[ADB100]

I would suspect either cabling or GBIC's. Check for errors on port 1/2.
Since you aren't using anything other than VLAN 1 why use a trunk?

Good luck

Andy

 

[omegaamigo]

I would suspect either cabling or GBIC's. Check for errors on port 1/2.

Yes, I'm beginning to think this myself. My suspicion is actually with the Cat5e cabling on the ports I've been testing 'West' from, rather than the fibre though.

Since you aren't using anything other than VLAN 1 why use a trunk?

Well, there are two switches and soon I will have more VLANs - i.e. when I introduce VMPS.

Thanks Andy...

 

[vipergg]

I see where one side is client but I couldn't see where you had set the other side as Server,client or transparent . If you have just 2 two switches you might just want to run them both as transparent . I would also prune off all vlans you are not using, "clear trunk 1/2 vlan 2-104,106-1005 " on both sides if you use transparent . I would also use something besides vlan 1 as this vlan carries all your control protocols like cdp , vtp etc... If the link appears slow then of course look at the physical layer stuff , make sure the correct fiber and gbics were used . If you used MM fiber with LX then you may need mode conditioning cables , you would have to look that up on the cisco site , don't know the specifics right off hand . Check the link on both sides and do a show counters 1/2 and look for link errors crc's etc... Usually slowness can be traced to physical layer stuff . And last looking at the config the ports are set to auto so make sure the client pc's are also set to auto otherwise you will have a speed/duplex mismatch .